A New Way To Estimate The Quality Of Cooling System Of LEDs

Last month, I published a post, Things About LED Cooling System. It said, the most direct method to estimate the quality of radiator of a LED lamp is to measure the temperature of LED chip when the lamp is working, the technical term of it is Junction Temperature. But obviously, this is a theory, Junction Temperature is hidden after a lamp is done working, this is the reason why we have to summarize so many methods to help us to estimate the quality of radiator of a LED lamp.

Today I'm gonna introduce anther way to measure the Junction Temperature indirectly.

"Illumination Temperature In Half An Hour" to test Junction Temperature

If we can not have the Junction Temperature in a direct way, how about a indirect way? The good news is, as a general rule, luminous flux will be decline while Junction Temperature is getting high. So, we just need to test and write down the change of Illumination Temperature of a lamp in a same spot, then we will get the Junction Temperature by working backward from it.

Specific practice is as follows:

1. Choose a occasion that can not be interfered by outside light. Night is a good choice, and turn off all other lights.
2. Turn on the lamp in cold state, test a Illumination Temperature of one spot, write down this number as Cold State Illumination Temperature.
3. Do not move the lamp and illuminometer's position, keep the lamp working.
4. Test Illumination Temperature of this same spot, write down this number as Hot State Illumination Temperature after half an hour.
5. If these two numbers tend to be similar(differ by 10~15%), this lamp’s cooling system is good.
6. If there is a great deal of difference between these two number(>20%), this lamp’s cooling system must be doubted.

I assume designers or Engineering contractors shall have an illuminometer? What!? You don't? Then go get one! It's not expensive at all.

However, what is the reasonable decline percentage of specific illumination temperature? Regarding this question, we have to know specific LED chip instead of stipulating the range simply.

Application of “Illumination Temperature In Half An Hour”

Here attached some "Luminous Flux VS Junction Temperature" changing curves of common chips.

• OSRAM S5 (3030) chip, its Junction Temperature exceeds 120℃ while luminous flux falls 20%.

• OSRAM S8 (5050) chip, its Junction Temperature exceeds 120℃ while luminous flux falls 20%.

• OSRAM E5 (5630) chip, its Junction Temperature exceeds 140℃ while luminous flux falls 20%.

• OSLON SSL80 White chip, its Junction Temperature exceeds 120℃ while luminous flux falls 15%.

From these pictures above, we could found that, if the Hot State Illumination Temperature is lower than the cold one by 20% after half an hour, then, basically, the Junction Temperature is out of tolerance range of chip. It means the cooling system failed.

This is most of the situation, should there be exceptions? Sure! At the end of the article, I'm gonna make an anti-example, you know, Glass half full.

Chip of OSLON Square 1-5W, luminous flux goes up when Junction Temperature rises from 25℃ to 40℃! Luminous flux starts to decrease After that. But, Luminous flux decreases 10% only while Junction Temperature rises to 120℃.

Summary:

• Heat conduction and Heat dissipation is absolutely important to LEDs.
• The key point is the Junction Temperature.
• You don't feel hot when you’re touching the radiator, it doesn't mean the radiator is definitely in good quality.
• If you feel hot when you’re touching the radiator, it's bad definitely.
• The suggestion would be Using The Method Of "Illumination Temperature In Half An Hour".
• Temperature curve of Chip determines the allowable variation range.